A molecular biological approach will be used to study fundamental relationships between the structure of the light (NF-L), mid-sized (NF-M) and heavy (NF-H) neurofilament (NF) polypeptide subunits and important features determining neurofilament function within the cell, namely those of NF assembly and NF phosphorylation. Initial studies will focus on the elements of NF subunit structure that are necessary and sufficient for association between subunits and their assembly into filaments. Comparative studies will examine the process of association and assembly between purified subunits obtained in highly phosphorylated forms from bovine and human tissues and between native subunits that are expressed and purified from bacterial systems. The latter studies will be conducted on products of E. coli plasmid expression vectors containing DNA encoding full-length, truncated or modified forms of NF-L, NF-M and NF-H. Particular attention will be directed at constructing altered subunits that represent only the "rod" and NH2-terminal domains, since these are the regions of NF polypeptides that are most likely required for filament assembly. Dimer, tetramer and filament formation between individual and admixed subunits will be monitored by biochemical, biophysical and morphological criteria, including the use of chemical cross-linking methods, "affinity" blots and electron microscopy. Subunit interactions and assembly will also be tested in a complex cellular system by introducing genomic DNA for NF-L, NF-M and NF-H, either singly or in combinations, into permissive mammalian cell lines and examining the states of assembly and phosphorylation of NF proteins which are expressed. A similar molecular approach will be used to characterize the highly-charged, COOH-terminal domains of NF-M and NF-H as well as the nature of their multiphosphorylation sites. COOH-terminal domains, purified from tissues and from bacterial systems, will be used to study their associations with one other, with filament-forming domains of NF polypeptides and with other axonal proteins. Multiphosphorylation domains will also be used to identify and characterize NF kinase(s) of neural tissues. Long-range studies (-04 and -05 yrs) will introduce human NF genes into transgenic mice in order to elucidate relationships between gene expression and the complex post-translational patterns of NF metabolism, especially those of NF assembly and of NF phosphorylation.